In mathematics , especially in the theory of Lie algebras , Cartan subalgebras are used, among other things, in the classification of semi-simple Lie algebras and in the theory of symmetric spaces . The rank of a Lie algebra (or its associated Lie group ) is defined as the dimension of the Cartan subalgebra. An example of a Cartan sub-algebra is the algebra of diagonal matrices.
definition
Let it be a Lie algebra. A subalgebra is a Cartan subalgebra if it is nilpotent and self-normalizing , that is, if
G
{\ displaystyle {\ mathfrak {g}}}
a
⊂
G
{\ displaystyle {\ mathfrak {a}} \ subset {\ mathfrak {g}}}
[
a
,
[
a
,
[
⋯
,
[
a
,
a
⏟
n
]
⋯
]
]
]
=
0
{\ displaystyle \ underbrace {[{\ mathfrak {a}}, [{\ mathfrak {a}}, [\ dotsb, [{\ mathfrak {a}}, {\ mathfrak {a}}} _ {n}] \ dotsb]]] = 0}
for one and
n
∈
N
{\ displaystyle n \ in \ mathbb {N}}
∀
Y
∉
a
∃
X
∈
a
:
[
X
,
Y
]
∉
a
{\ displaystyle \ forall Y \ not \ in {\ mathfrak {a}} \ \ exists X \ in {\ mathfrak {a}}: \ \ left [X, Y \ right] \ not \ in {\ mathfrak {a }}}
applies.
Examples
A Cartan subalgebra of
G
=
s
l
(
n
,
C.
)
=
{
A.
∈
M.
a
t
(
n
,
C.
)
:
S.
p
u
r
(
A.
)
=
0
}
{\ displaystyle {\ mathfrak {g}} = {\ mathfrak {sl}} (n, \ mathbb {C}) = \ left \ {A \ in \ mathrm {Mat} (n, \ mathbb {C}): \ mathrm {track} (A) = 0 \ right \}}
is the algebra of the diagonal matrices
a
0
=
{
d
i
a
G
(
λ
1
,
...
,
λ
n
)
:
λ
1
+
...
+
λ
n
=
0
}
{\ displaystyle {\ mathfrak {a}} _ {0} = \ left \ {\ mathrm {diag} (\ lambda _ {1}, \ ldots, \ lambda _ {n}): \ lambda _ {1} + \ ldots + \ lambda _ {n} = 0 \ right \}}
.
Every Cartan sub-algebra is too conjugate .
a
⊂
s
l
(
n
,
C.
)
{\ displaystyle {\ mathfrak {a}} \ subset {\ mathfrak {sl}} (n, \ mathbb {C})}
a
0
{\ displaystyle {\ mathfrak {a}} _ {0}}
In contrast, Cartan has two unconjugated subalgebras, viz
s
l
(
2
,
R.
)
{\ displaystyle {\ mathfrak {sl}} (2, \ mathbb {R})}
a
1
=
R.
(
1
0
0
-
1
)
{\ displaystyle {\ mathfrak {a}} _ {1} = \ mathbb {R} \ left ({\ begin {array} {cc} 1 & 0 \\ 0 & -1 \ end {array}} \ right)}
and
a
2
=
R.
(
0
1
1
0
)
{\ displaystyle {\ mathfrak {a}} _ {2} = \ mathbb {R} \ left ({\ begin {array} {cc} 0 & 1 \\ 1 & 0 \ end {array}} \ right)}
.
Existence and uniqueness
A finite-dimensional Lie algebra over an infinite field always has a Cartan subalgebra.
For a finite-dimensional Lie algebra over a field with characteristic it holds that all Cartan subalgebras have the same dimension.
0
{\ displaystyle 0}
For a finite-dimensional Lie algebra over an algebraically closed field , all Cartan subalgebras are conjugated to one another , namely under the group that is generated by the automorphisms (for in Lie algebra and nilpotent).
exp
(
a
d
(
X
)
)
{\ displaystyle \ exp (\ mathrm {ad} (X))}
X
{\ displaystyle X}
a
d
(
X
)
{\ displaystyle \ mathrm {ad} (X)}
properties
If a semi-simple Lie algebra is over an algebraically closed field , then every Cartan sub -algebra is Abelian and the restriction of the adjoint representation to is simultaneously diagonalizable with as eigenspace to weight . That is, there is a decomposition
G
{\ displaystyle {\ mathfrak {g}}}
a
⊂
G
{\ displaystyle {\ mathfrak {a}} \ subset {\ mathfrak {g}}}
a
d
:
G
→
G
l
(
G
)
{\ displaystyle \ mathrm {ad} \ colon {\ mathfrak {g}} \ to {\ mathfrak {gl}} ({\ mathfrak {g}})}
a
{\ displaystyle {\ mathfrak {a}}}
a
{\ displaystyle {\ mathfrak {a}}}
0
{\ displaystyle 0}
G
=
a
⊕
⨁
α
∈
a
∗
G
α
{\ displaystyle {\ mathfrak {g}} = {\ mathfrak {a}} \ oplus \ bigoplus _ {\ alpha \ in {\ mathfrak {a}} ^ {*}} {\ mathfrak {g}} _ {\ alpha}}
With
a
d
(
X
)
(
Y
)
=
[
X
,
Y
]
=
α
(
X
)
Y
∀
X
∈
a
,
Y
∈
G
α
{\ displaystyle \ mathrm {ad} (X) (Y) = \ left [X, Y \ right] = \ alpha (X) Y \ quad \ forall X \ in {\ mathfrak {a}}, Y \ in { \ mathfrak {g}} _ {\ alpha}}
and
G
α
≠
0
⟹
α
(
X
)
≠
0
∀
X
∈
a
{\ displaystyle {\ mathfrak {g}} _ {\ alpha} \ not = 0 \ Longrightarrow \ alpha (X) \ not = 0 \ quad \ forall X \ in {\ mathfrak {a}}}
.
For example
G
=
s
l
(
n
,
C.
)
=
{
A.
∈
M.
a
t
(
n
,
C.
)
:
S.
p
u
r
(
A.
)
=
0
}
,
{\ displaystyle {\ mathfrak {g}} = {\ mathfrak {sl}} (n, \ mathbb {C}) = \ left \ {A \ in \ mathrm {Mat} (n, \ mathbb {C}): \ mathrm {track} (A) = 0 \ right \},}
a
=
{
d
i
a
G
(
λ
1
,
...
,
λ
n
)
:
λ
1
+
...
+
λ
n
=
0
}
{\ displaystyle {\ mathfrak {a}} = \ left \ {\ mathrm {diag} (\ lambda _ {1}, \ ldots, \ lambda _ {n}): \ lambda _ {1} + \ ldots + \ lambda _ {n} = 0 \ right \}}
is when the elementary matrix is designated with entry at that point and entries otherwise
e
i
j
{\ displaystyle e_ {ij}}
1
{\ displaystyle 1}
(
i
,
j
)
{\ displaystyle (i, j)}
0
{\ displaystyle 0}
G
=
a
⊕
⨁
i
≠
j
C.
e
i
j
=
a
⊕
⨁
α
∈
a
∗
G
α
{\ displaystyle {\ mathfrak {g}} = {\ mathfrak {a}} \ oplus \ bigoplus _ {i \ not = j} \ mathbb {C} e_ {ij} = {\ mathfrak {a}} \ oplus \ bigoplus _ {\ alpha \ in {\ mathfrak {a}} ^ {*}} {\ mathfrak {g}} _ {\ alpha}}
with for
C.
e
i
j
=
G
α
{\ displaystyle \ mathbb {C} e_ {ij} = {\ mathfrak {g}} _ {\ alpha}}
α
(
λ
1
,
...
,
λ
n
)
=
λ
i
-
λ
j
{\ displaystyle \ alpha (\ lambda _ {1}, \ ldots, \ lambda _ {n}) = \ lambda _ {i} - \ lambda _ {j}}
.
literature
Élie Cartan : Sur la structure des groupes de transformations finis et continus. Thèse, Paris 1894.
Anthony W. Knapp: Lie groups beyond an introduction. (Progress in Mathematics, 140). Second edition. Birkhäuser, Boston, MA 2002, ISBN 0-8176-4259-5 .
<img src="https://de.wikipedia.org/wiki/Special:CentralAutoLogin/start?type=1x1" alt="" title="" width="1" height="1" style="border: none; position: absolute;">